So I'm given a TFile that contains two TTree objects, which contain track/tower pT, eta and phi, divided by events. My goals is to extract each track and tower in the event and then cluster whole event using the FastJet package. Now, if I'm doing this task using pure ROOT my analysis takes 30 minutes at max (with ~100 GB TFile). In the meanwhile, uproot will process only 10,000 events in this time limit...
It is apparent that I'm doing something wrongly, so I wanted to ask, what would be the proper way to access track-by-track information to get the same speed as in ROOT?